Methods for controlling access to data via computer networks are well known. Some methods require that IDs and passwords be provided to the data server in order to gain access to data. In other methods a data server provides encrypted data to the client, where the data is then decrypted using a key and decryption software previously provided to the client. In other methods digital watermarks are embedded into the data at the data source allowing for distribution of the data to be monitored for unauthorized distribution.
These and other prior art approaches are suited for peer-to-peer architectures where the data resides at a single source, such as on one or more servers under the direct control of the data owner, and the data is provided directly from the data holder's server to the requesting client. Such peer-to-peer arrangements are currently in wide use on the Internet, and represents the most common default arrangement.
Providing high-bandwidth content via computer networks, however, places a considerable load on peer-to-peer architectures. As a result proxy servers are often used to “mirror” data stored at the data owner's server and provide data to clients. Unfortunately, these proxy servers are often not under the direct control of the data owner, and steps taken by the data owner at the original data source to protect unauthorized access to data are often not taken by the proxy servers, leading to unauthorized access.